Coatings of Polymers and Plastics Part 5 ppt

Formulating Plastics for Paint Adhesion 89 γ S —the surface energy of the solid (the energy necessary to increase the surface of the solid) γ L —the surface energy of the liquid (the energy necessary to increase the surface of the liquid) which is valid at equilibrium. Figure 1 shows a representation of the physical situation. When θ is zero, cos θ is 1, and wetting occurs with the surface energy of the solid being equal to or greater than the liquid, depending on the interfacial energy. The surface energies of the solid, the liquid, and the interface are mate- rial properties; the contact angle is measured. One can see that liquids with low surface energy will wet out onto solids with high surface energy, because the vector force is to “pull the liquid down.” This means that oil (low surface en- ergy) spreads out on water (high surface energy), but water (paint in this case), doesn’t spread out on oil (“solid oil” like polyolefins). For a given γ SL , interfacial energy, a high surface-energy solid is necessary because the surface energy of the liquid is reduced by the cos θ, which varies from 0 to 1. The interfacial surface energy is an important component of this equation, but can not be measured directly (9). The surface energy of the solid can be determined by extrapolation from liquid homologs (8) or by wettability data with various surface-tension liquids. A theoretical model is needed to relate the interfacial energy to the surface energy of the two components. Good and Girifalco developed a very early model; however, this model did not include the fact that the surface energy has two components: a dispersion component, γ d , and a polar component, γ p , for each material. The term dispersion comes from the fact that the perturbation of electronic motion that creates this force is related to the perturbation of light with frequency (or dispersion of light) (8). The adhesion between the two materials is described by the following equation where W AB is the work of adhesion, γ A , γ B , and γ AB are the surface F IG .1 Spreading of liquid onto solid and contact angle. 90 Berta energies of material A, material B, and the interfacial surface energy between A and B, respectively (8). WAB =γ A +γ B −γ AB Eq. (2) For a high work of adhesion (e.g., good paint adhesion) the materials are diffi- cult to separate, unless a strong enough force is applied to exceed the work of adhesion. It can be seen from Eq. (2) that a small interfacial surface energy (which is the energy necessary to create a surface between the A and B) would lead to greater work of adhesion. It should be stated here that when the interfa- cial energy is zero, the materials are thermodynamically miscible. It can also be seen that high surface-energy materials (typical of polar materials), also lead to high work of adhesion. Realizing that it is desirable to minimize the interfacial energy between the surface of the TPO (material A) and the paint surface energy (material B), we set out to do this utilizing the considerations below. All materials have both a polar (γ p A ) and nonpolar (γ d A ) or dispersion contri- bution to the total surface energy. There are two main models for determining the interfacial energy that consider both contributions to surface energetics for each material—the harmonic mean model and the geometric mean model (9). The harmonic mean model is as follows: γ AB =γ A +γ B − (4γ p A γ p B )/(γ p A +γ p B ) − (4γ d A γ d B )/(γ d A +γ d B ) Eq. (3) The geometric mean model is as follows: γ AB =γ A +γ B − 2(γ p A γ p B ) 1/2 − 2(γ d A γ d B ) 1/2 Eq. (4) To minimize the interfacial energy, the polar and nonpolar contributions for the two materials should match. To see this, consider material A has a 20/1 polar to nonpolar surface energy and material B has a 1/20 polar to nonpolar surface energy (total surface energy for each material is 21 erg/cm 2 ). The interfacial energy, γ AB , is 24 erg/cm 2 by Eq. (4). Contrast this with each material having a 10.5/10.5 polar to nonpolar surface energy, which would lead to a value of zero (0 erg/cm 2 ) for the interfacial energy, or a minimum of interfacial energy and a maximum for adhesion. It can been seen from Figure 2, which is a plot of interfacial energy for two materials, A and B (each of which has a total surface energy of 21 erg/cm 2 ) as a function of the polar surface energy of material A (or the dispersion surface energy of material B), that a match of dispersion and polar gives an interfacial energy minimum. For the sake of simplicity, we assumed a symmetrical bal- anced contribution for each material between the dispersion and polar contribu- tions to the total surface energy. Both the harmonic mean model and the geometric mean model will be used to determine the surface energy of the solid surface along with the polar Formulating Plastics for Paint Adhesion 91 F IG .2 Interfacial energy of AB material blend. and dispersion contributions for the compositions considered in the formulation section. 3 THE CHEMISTRY Chemistry is required for the paint to cure and bonds to form to provide the forces that are necessary to achieve the desired paint properties, and also to obtain the adhesion of the paint to the substrate or TPO (the material of focus). The classification of paint curing as a chemical change is obvious, just as the classification of a phase change—such as melting of a solid—as a physical change is also obvious; however, the lines get a little blurred when we start to include the interactions of strong polarity, hydrogen bonding, charge transfer, etc. Realizing that these all stem from one force of nature (the electromagnetic force); and that the other three forces (10) (gravity, the strong nuclear force, the weak nuclear force) are of no consideration here. Probably a good way to clas- sify the forces involved is the one used by Coleman (11), wherein at the lower end of the scale are the weak “physical” forces, generally referred to as van der Waals forces, in which the force is proportional to the inverse sixth power of the distance between the interacting species; and at the upper end, strong “chem- ical” forces, such as ionomers, charge transfer, and of course the classical “chemical reaction” such as acid-base, alcohol-acid, etc. Hydrogen bonding is classified by Coleman as a chemical change, and of intermediate strength. How- ever, as Wu points out, it is not a primary chemical bond and is mainly of ionic character by nature (8). This classification becomes important as we consider the chemical species involved in the process of paint adhesion and achieving good adhesive strength; they also become important in our selection process of materials to consider as key formulation factors. Thirdly, they become important as we try to develop a deeper understanding, and a working model to explain the results and build on them to achieve a final objective of a useful, commercial material. 92 Berta 4 STRENGTH OF BONDS 1. Physical Bonds a. Random dipole-induced dipole (London forces) E = f(ionization potential, polarizability)/r 6 Eq. (5) Strength ϳ 5 kcal/mol b. Dipole-induced dipole (Debye) E = f(dipole moment, polarizability)/ r 6 Eq. (6) Strength ϳ 5 kcal/mol c. Dipole-dipole (Keesom) E = f(dipole moment)/r 3 Eq. (7) or E = f(dipole moment, 1/T)/r 6 Eq. (8) Strength ϳ 8 kcal/mol d. Ion-dipole E = f(dipole moment, charge)/r 2 Eq. (9) Strength ϳ 5–10 kcal/mol 2. Chemical Bonds a. Hydrogen bonding E = f(acceptor / donor attraction)/r 6 Eq. (10) Strength ϳ 5 to 40 kcal/mol b. Covalent bonding Strength ϳ 20 to 200 kcal/mol c. Ionic bonding Strength ϳ 120 to 250 kcal/mol To exemplify how dramatically the energy drops off with distance for a van der Waals attraction (r −6 dependency): assume E = 20 kcal/mol for r and E = 0.3 kcal/mol for 2r This means that for a typical attraction with dimensions between centers of a few angstroms, the driving force of energy released due to bonding becomes very small just a short distance from the equilibrium bond distance. Thermal motions, steric hindrance, or segmental restrictions make disruption easy. For a lower order distance dependency, such as ion-dipole, the energy reduction at twice the distance is much less: assume E = 20 kcal/mol for r (for r −2 dependency) and E = 5 kcal/mol for 2r For a typical ionic bond (12) as the centers approach each other, they are sucked into the potential energy well. At a distance of r of 2.3 A, the bond energy is 122 kcal/mol and at almost twice the distance, 4 A, the energy is still very high at 82 kcal/mol. Covalent bonds behave in a similar fashion. Formulating Plastics for Paint Adhesion 93 Other key factors that are important are geometrical considerations (as in the case with hydrogen bonding) and spacing between the segments involved (which could make near distances forces difficult to realize). In the case of proteins, many hydrogen bonds are formed, which lock the structure into place. The paint chemistry must now be considered. The details of all the paint chemistries involved can be found in several good books on coatings (13,14). To simplify the situation, we will mention the most prominently used systems, which are the ones involved in the development of directly paintable TPOs, and which have been the ones most commonly used by the automotive industry for painting TPOs. The urethane chemistry used for paint involves isocyanate groups and hydroxyl groups, the reaction of which forms a urethane. Of course, there are catalysts included and a great deal of formulation work with the ingre- dients involving various molecular weights and chemically functional groups, but what is important for our purpose is the main reactions or functionality that would guide in the selection of additives for TPO. Figure 3 shows a representa- tion of the isocyanate-hydroxy reaction to form a urethane. The isocyanate can also react with amines to form ureas (just replace the O with an N in the struc- ture, but don’t leave out the extra hydrogen). Hydroxy-terminated polymers are good materials to consider for addition to TPO. The urethane paints can usually be cured at lower temperature, such as 80°C. The other predominant chemistry for curing paints involves melamine with ether groups that can react with hydroxyl groups of polymeric paint addi- tives to cure (increase the mole weight) by transetherification with the low mole weight hydroxy-material from the melamine evaporating off (see Fig. 3) (13). Ordinarily curing is done at higher temperatures, for example 121°C. Once again, hydroxy-functionality is useful. Figures 4 and 5 show representations of hydrogen bond formation between an electron donating group (such as carbonyl or ether) and a hydrogen attached to an electron withdrawing group (such as halogen or carboxylic acid) that makes the hydrogen more available to bond. 5 THE FORMULATIONS 5.1 Basic Considerations This section now will discuss work and formulations, some of which have al- ready been disclosed in the literature, and the approach used herein to develop a balance of paint adhesion, durability, and good physical properties. Mainly Clark and Ryntz (15,16), and others (17), have worked on the development of directly paintable TPO using amine-terminated polyethylene oxide (ATPEO) reacted with maleic anhydride grafted polypropylene to form an imide. This allows for a tie-in to the PP matrix and the functionality (presumably through 94 Berta F IG .3 Paint curing chemistry. hydrogen bonding of the ethers groups to hydrogen donating groups in the paint, such as hydroxyls) needed to provide paint adhesion. Well-defined compositions with a emphasis on a favorable balance of the level and ratio of the MAgPP and ATPEO were explored. Their work (15,16) is a foundation for continued extension and improvements. The literature (18,19) also shows the use of hy- droxyl functionality on PP to improve adhesion to paint. In addition, polyesters Formulating Plastics for Paint Adhesion 95 F IG .4 Hydrogen bonding with halogen substitution and a carbonyl group. have been added (20) to improve the paint adhesion. A combination of MAgPP and an epoxy resin has also been used (21,22). Other work has been published on improved coatability of TPOs (23,24). By and large, adhesion is addressed with various paints and curing conditions, and various tapes are used to do the testing. In most cases, multiple pulls were not addressed, and the durability was usually not tested or exemplified. This may be due to the fact that both adhesion and durability are difficult to balance. It is difficult to get a good balance of adhesion and durability in a normal TPO with an adhesion promoter (25). The problem is magnified in DPTPO by the necessity of purposely having to add a polar material to the TPO. With the formulations described in the tables and the following text, we will be addressing both adhesion and durability, using a very aggressive adhesion test that involves multiple pulls with a very good adhesive tape that sticks well to the paint (this is critical to a good test) (D. Frazier, formerly of Montell Polyolefins, private communication). The effects of shear on the adhesion results are also evaluated by employing a specific test (uncov- ered during this work) that is simple yet surprisingly very effective. A concep- tual model has been developed to address both these problems and has been utilized to formulate the essential requirements for a commercially acceptable DPTPO. F IG .5 Hydrogen bonding with an ether group and a hydrogen of a carboxyl group. 96 Berta From an surface energetics viewpoint, referring back to the section on the theory of adhesion and Eq. (4), the aim is to match the surface energy of the DPTPO with the surface energy of the paint. Table 1 shows the surface energet- ics of three materials; polyolefin, polyvinylchloride (PVC), and a RIM polyure- thane (PU); and two coating materials: adhesion promoter (containing chlori- nated polypropylene) and a melamine paint, which represents quite nicely about the average surface energy of paints used for TPO with only about a few points difference from paint to paint. It is easy to see why PVC paints well, because it matches the surface energetics of paint with bonding forces being of the weaker type, which probably involve hydrogen bonding of hydroxyl groups with chlo- rine. The RIM, most probably, involves some chemical reaction of isocyanate gropus in the paint with the hydrogen on the nitrogen group of the RIM ure- thane; also hydroxyl groups in the paint would form hydrogen bonds with the carbonyl of the RIM urethane. An adhesion promoter would “bridge the gap” in terms of surface energetics between the paint and the TPO, acting as a tie- layer. The target surface energetics for a DPTPO would be to equal, as nearly as possible, the paint surface energetics. One must realize that this energetics match would not guarantee strong, durable bonding of the paint to the TPO, and that the near-surface and deeper-surface effects based on “compatibility” of additives are also critical. By utilizing several polar ingredients of various degrees of polarity, we have been able not only to effect a better balance of paint adhesion and durabil- ity, but also to minimize the of bulk property effects due to the incompatibility of the polar additives. For example, a multicomponent polarity balanced distri- bution (MCPBD) that utilizes (1) a highly modified propylene polymer, (2) a moderately modified propylene polymer, (3) a polar additive capable of reactiv- ity, and (4) an interfacial modifier of moderately low polarity was developed and has shown good success, not only in the lab, but also when scaled up to commercial size trials. The first grade of DPTPO developed was a low modulus (650 MPa) using this MCPBD model approach. In the development of such materials, it is necessary also to consider the influence of shear forces during molding, on the surface and near surface properties; this will be dealt with later on in this section. T ABLE 1 Surface Energy Matches Surface energy Adhesion (ERG/cm 2 ) PO PVC RIM promoter Paint Total 29.8 38 21.7 44.2 42 Dispersion 27.6 31.6 13.7 38.4 31.5 Polar 2.2 6.4 8 5.8 10.5 Formulating Plastics for Paint Adhesion 97 5.2 Compression Molded Level As stated previously, notwithstanding this challenge, significant progress has been made by Richard Clark (of Luzenac, formerly of Texaco, private communi- cation) and Rose Ryntz toward the development of a directly paintable TPO, which has contributed significantly to the understanding of such systems and directly paintable TPO in general. We have taken a somewhat similar fundamen- tal approach using reactive functional ingredients, but not just two. By incorpo- rating several other ingredients into the mix, the result was a more advantageous balance of material properties and paint performance with favorable cost consid- erations. Table 2 shows the compositions made by compression molding along with the paint adhesion and surface energetics results. In this case, the MAgPP-1 (maleic anhydride grafted PP) used was made by grafting onto the surface of the solid phase. Although the MAgPP-1 shows some polarity (T 2-1) compared to PP (T 2-8), or a PP-EPR blend (T 2-7), it shows no adhesion. This is probably T ABLE 2 Effect of Functionalized Polyolefins a Composition T 2-1 T 2-2 T 2-3 T 2-4 T 2-5 T 2-6 T 2-7 T 2-8 MAgPP-1 100 14 14 — 14 — — — PP — 56 56 70 56 70 70 100 MAgEPR-1 — — — 15 15 — — — EPR-1 — 30 30 15 15 30 30 — ATPEO-1 — — 6 — — 6 — — Paint adhesion (% adhesion) 1st pull 0 100 100 10 100 0 0 0 2nd pull — 100 100 0 100 — — — 3rd pull — 100 100 — 100 — — — Surface energy WORK (Erg/cm 2 ) model Total 29.7 37.7 49.6 23.4 36.4 30 28.9 28.6 Dispersion 20.7 35.7 23.3 18.5 29 24.8 26.6 26.5 Polar 9 2 26.3 4.8 7.4 5.2 2.3 2.1 Surface energy Wu (Erg/cm 2 ) model Total — 38 52.3 28 39.6 33.3 31.3 30.7 Dispersion — 31.3 20 19.1 27.4 23.3 24.6 24.9 Polar — 6.6 32.3 8.9 12.2 10 6.5 5.8 a Compression molded samples, HAP-9440 paint. 98 Berta due to the inability to obtain significant diffusion into the molded part (poor diffusion across the crystalline phase). With about 30% rubber, PP, and MAgPP-1, good adhesion is realized with (T 2-3) or without (T 2-2) the amine terminated polyethylene oxide (ATPEO); the former does have very high surface polarity with the latter having very high dispersion energetics. Compare T 2-2 and T 2-3. If a maleic anhydride EPR is used, MAgEPR-1 (T 2-4), somewhat higher polar- ity is achieved, but the dispersion component is low and the adhesion is poor. Using both the MAgPP-1 and the MAgEPR-1, both polarity and dispersion com- ponents are high, and adhesion is good (T 2-5). Based upon adhesion only, it is not necessary to combine the two (MAgPP and MAgEPR), but we will see later that this will become important as we delve into durability, effects of shear, and physical properties. It turns out that the surface energetics of T 2-5 just about match the surface energetics of a typical paint HAP-9940, which is about 30 erg/cm 2 for dispersion, 10 erg/cm 2 for polar, and, obviously, 40 erg/cm 2 total. It will also be noted that just by adding the ATPEO-1 (T 2-6) polarity is increased, but there is not adhesion. Probably because the ATPEO comes to the surface, but has no tie-in. A similar result would also be expected with the addition of various slip additives or surfactants. Both the WORK (Wendt, Owens, Rabel, and Keabele) model and the Wu model were used to determine surface energet- ics, with the latter giving higher polarity for a simple polyolefin than what is generally accepted in the literature. Therefore, the former or WORK model is to be preferred, in the context of this work. Table 3 shows results with good adhesion and surface energy when using another type of MAgPP-2 made in a molten process (T 3-1). This MAgPP supplied by Eastman was deemed to be the preferred material for directly paint- able TPO (R. Clark, private communication). Again, the ATPEO-1 is not neces- sary to obtain adhesion for the compression-molded plaques, but certainly does enhance the surface energetics (T 3-2). This becomes important as we move to a better understanding of what is necessary to achieve DPTPO. Formulation T 3-3 with MAgPP added to PP and no rubber phase, both adhesion and surface energetics are poor due to the difficulty of diffusion into what is solely a crystal- line phase. The compression molding process is quite different than the injection molding process, as will be seen later, and as others have found. Reyes, et al. (26) have shown good results with compression molding, but virtually not adhe- sion with injection molded samples of the exact same composition. It is impor- tant to note that the availability of sufficient maleic anhydride functionality is all that is needed to obtain adhesion. This functionality reacts with the functional groups in the paint system (which in this case was melamine type). It is also important to note that the MAgPP has to be available on or near the surface, which it apparently is in the case of compression molded samples. For a reactor TPO (RTPO-1), the results (T 3-4 and T 3-5) are the same as for a compounded blend of PP and EPR. [...]... better choice (T 5- 3 versus T 5- 6) and (2) the ATPEO is necessary along with the MAgPP to realize adhesion for injection-molded parts As the amount of MAgPP and ATPEO increase, better adhesion is achieved (T 5- 2 to T 5- 4) along with good surface energetics; Formulating Plastics for Paint Adhesion 101 TABLE 5 Adhesion Durability Imbalancea Composition T 5- 1 T 5- 2 T 5- 3 T 5- 4 T 5- 5 T 5- 6 T 5- 7 RTPO-1 MAgPP-2... ATPEO-2 100 — — — — 100 12 .5 — — 2 100 20 — — 4 100 20 — — 6 100 — 12 .5 — 2 100 — 20 — 4 100 20 — 20 4 Paint adhesion (% adhesion) 1st pull 2nd pull 3rd pull 4th pull 0 — — — 75 45 — — 90 90 80 45 100 100 100 100 0 — — — 0 — — — 100 100 100 100 Durability (% failure) 50 cycles 100 cycles 0 0 5 20 33 45 55 90 45 90 25 33 0 0 33 .5 28.3 5. 2 36.8 28 8.8 39.6 27 12.6 28 25. 4 2 .5 28 .5 26.4 2.1 Surface energy... pull Durability (% failure) 50 cycles 100 cycles T 11-1 T 11-2 T 11-3 T 11-4 T 11 -5 T 11-6 100 — 10 5 — 10 10 3 — 100 — 10 5 5 10 15 — 3 100 — 10 5 5 10 20 — 2 100 — 12 5 5 12 30 — — 100 — 100 10 10 — 10 — — 3 Gate/opp 98/100 76/100 68/100 56 /00 12 10 5 12 50 — — Gate/opp Gate/opp Gate/opp Gate/opp Gate/opp 100/100 100/100 100/100 100/100 40/ 65 100/100 100/100 96/100 96/100 30 /50 100/100 98/100 74/100... thrust of this chapter Once again it should be Formulating Plastics for Paint Adhesion 109 TABLE 10 Effect of Cure Conditionsa Cure time, (min.) 20 Paint adhesion (% adhesion) 1st pull 2nd pull 3rd pull 4th pull 5th pull a 30 40 50 Gate/opp 80/ 85 70/ 65 55/ 40 40/30 28/ 25 Gate/opp 100/100 92/100 90/100 70/98 60/88 Gate/opp 100/100 100/100 100/100 100/100 100/100 Gate/opp 95/ 100 90/100 88/100 85/ 100 80/100... — 100 100 100 100 50 0 hours NA NA 100 100 100 100 1,400 hours NA Xenon arc, center of disc tested NA 95 88 73 64 112 Berta TABLE 13 Conductive DPTPOa Composition T 13-1 T 13-2 T 13-3 T 13-4 T 13 -5 100 10 5 5 10 10 3 — — 100 10 5 5 10 10 3 2 — 100 10 5 5 10 10 3 4 — 100 10 5 5 10 10 3 — 2 100 10 5 5 10 10 3 — 4 Paint adhesion (% adhesion) 1st pull 2nd pull 3rd pull 4th pull Gate/opp 100/100 100/100... Adhesion 1 05 FIG 8 Better additive system balance—no delamination material and parts were molded on commercial-size injection-molding equipment The first grade of DPTPO developed was a low modulus ( 650 MPa) using this MCPBD model approach Typical automotive parts were made and tested for paint adhesion and durability with acceptable results (R.A Ryntz, private communication) 5. 5 Paint Type and Curing... 12 -5 100 10 5 7 10 10 3 — — — 100 10 10 — 10 10 3 0.11 — — 100 10 10 — 10 10 3 — — 0.14 100 10 10 — 10 10 3 — 1 — 100 10 5 7 10 15 3 — 2 — 100 100 100 100 100 100 100 95 80 75 100 100 100 100 93 82 60 22 100 100 100 100 53 23 — — 100 100 100 100 0 hours 100 100 100 100 100 100 100 100 100 200 hours 50 20 — — 70 50 — — 100 100 100 100 50 0 hours NA NA 100 100 100 100 1,400 hours NA Xenon arc, center of. .. 7 Effect of Tape Used on Adhesion Resultsa DPTPO-1 (T 5- 4) RTPO-1 MAgPP-2 ATPEO-2 Paint adhesion (% adhesion) Type of tape 1st pull 2nd pull 3rd pull 4th pull 5th pull a DPTPO-2 (T 5- 2) 100 20 6 100 20 6 100 20 6 100 12 .5 2 100 12 .5 2 100 12 .5 2 Masking 100 100 100 100 100 Scotch 100 100 100 100 100 3M 898 100 100 100 100 100 Masking 100 100 100 100 100 Scotch 100 100 98 85 80 3M 898 75 45 — — — Injection... Gate/opp Gate/opp 100/100 100/100 100/100 100/100 40/ 65 100/100 100/100 96/100 96/100 30 /50 100/100 98/100 74/100 80/100 — 100/100 94/100 50 /100 50 /100 — 0 0 0 0 2 8 3 12 6 14 85 95 Flexural 1% tan (MPa) 650 7 25 8 65 1 050 1 450 1800 Izod impact −20 C (ft-lbs) 11.3 2.7 1.2 0.9 0. 85 — 42 36 34 22 29 — Ceast impact −30 C (J) a Injection molded discs, DuPont 872 paint, Hot Taber Durability This property has been... 100 5 10 10 10 — Gate/opp 94/100 75/ 100 70/100 44/76 Gate/opp 98/100 74/100 50 /98 46/74 Pain adhesion (% adhesion) 1st pull 2nd pull 3rd pull 4th pull Gate/opp 72/100 68/100 52 /100 42/98 Durability (% failure) 50 cycles 100 cycles 0 8 0 0 80 90 90 85 0 0 0 0 Dynamic impact (J) −30 C −40 C 32 16 10 2 35 28 36 26 24 5 38 27 Izod impact −20 C (ft-lbs/in) 2.7 1.3 2.6 2.3 1.7 2.6 Flexural modulus (MPa) 55 0 . (T 5- 2 to T 5- 4) along with good surface energetics; Formulating Plastics for Paint Adhesion 101 T ABLE 5 Adhesion Durability Imbalance a Composition T 5- 1 T 5- 2 T 5- 3 T 5- 4 T 5- 5 T 5- 6 T 5- 7 RTPO-1. 45 100 — — 100 Durability (% failure) 50 cycles 0 5 33 55 45 25 0 100 cycles 0 20 45 90 90 33 0 Surface energy (Egr/cm 2 ) Total 25. 6 33 .5 36.8 39.6 28 28 .5 37.9 Dispersion 22.6 28.3 28 27 25. 4. 100/100 10/100 15/ 100 Durability (% failure) 50 cycles 10 35 25 0 0 0 45 0 100 cycles 15 50 35 0 0 0 70 0 a Injection molded discs, DuPont 872 paint, Hot Taber Durability. Formulating Plastics for